mradermacher/DeepSeek-V2-Lite-FP8-BLOCK-Fused-i1-GGUF

Quantization vs nm-testing/DeepSeek-V2-Lite-FP8-BLOCK-Fused. nm-testing/DeepSeek-V2-Lite-FP8-BLOCK-Fused is a quantized version of the other (F8_E4M3/BF16).

On the Hub · Downloads on Model Pulse